Restrictive, insofar as it's places more restrictions on you than a work that is not protected by copyright. GPL relaxes the restrictions that copyright places on you. It does not eliminate them or add new restrictions. It does trade absolute restrictions (Thou shalt not duplicate) for conditional restrictions (Thou shalt not duplicate, unless you also provide source under GPL).
So, to recap, the GPL is a restrictive license to the extent that relaxing a restriction without eliminating completely is itself a restriction.
Relying on slashdot to get washing machine mod ideas? Ack no! All you'll get are neon light inserts, acrylic panels, and maybe a Matrix Orbital display w/ backlit keypad to replace the dials on the top.
How would theming your washing machine help brew beer?
GPL says that any redistribution of a GPL'd work or derivative must include an offer of source. It doesn't say you need to make binaries available or installable ISOs available.
It's probably more like somebody happened to be looking for the ROM for some other purpose, noticed a bunch of symbols that looked familiar, and started poking around the Linux kernel to compare. (They may have even seen "Linux" or similar strings.) In this case, the affected code is PCI setup code, so maybe the person who noticed was debugging some problem on their particular system that uses this product, or were developing a PCI driver.
Actually, the odds are much better than that. Columbia flew well over 100 missions before we lost it, and we've had only two shuttle disasters. I seem to recall hearing that the odds are estimated at about 1 in 250 that a given shuttle launch will result in loss of equipment or personnel. Still not great odds, but nearly an order of magnitude better than 1 in 50.
What gets me is that even if NASA knew reentry was going to destroy the shuttle after reviewing the launch videos, there wasn't much (anything?) they could really do to save the astronauts. Rather, they get to float around for 2 weeks waiting to die. That's really neat.:-P
Threaded trees are kinda cool, but they do have some additional storage associated with them. Specifically, you have to have some way of distinguishing between a pointer to a child and a pointer back up into the tree. Also, how do you manage to get a post-order traversal out of one?
Back to the topic: Once you make the translation from code-recursion to data-recursion, you also open yourself up to other neat transformations. For instance, the difference between a depth-first search and a breadth-first search is simply the difference between a stack and a queue. If you write your code appropriately, you can switch between one or the other painlessly. And if you use a priority queue, you can have a loop which morphs between the two or gives you something in-between, just by changing the priority function. (I find this last structure useful for maze generation.)
So you trade code recursive for data recursive. It's still recursive. Often, data recursive code is more complex to get right, too.
As the parent post noted, some algorithms are inherently recursive. Unfortunately, many
teaching examples given to beginners don't need recursion.
Iterative tasks can be made recursive, but there's little point. For example, consider the classic "factorial()" and "fibonacci()" examples, which are really just iterative examples in which previous iterations feed future iterations. Although these tasks are "recursive" in the mathematical sense -- later elements are defined in terms of previous elements -- the computation itself is iterative.
Tasks which "fork", such as depth-first traversals of complicated structures, aren't simple iterative processes, since there is an implicit "go back" step that rewinds to a previous context where the fork occurs. You can translate these into a loop that lacks recursive function calls, but only if you sprout a data stack to keep track of these forking points. You've really just traded one stack for another and have not transformed the problem.
Why can't you remotely log in? Why can't you click a sequence of coordinates on an imagemap on a web page? The images, and their reactions to being clicked, need not reflect their occult nature.
And how does that help me log in with SSH, SCP, telnet or FTP? The webpage would need to give me some sort of textual authentication token as part of mastering a series of clicks. I would then give telnet/ftp/ssh/scp the textual authentication token in order to log in.
And where would I provide that token to these programs? At the password prompt.
I guess one thing you could do with this is make the passwords single-use only.
I suppose technically so. A slight change in wording is all that's necessary. The GPL requires you (see paragraph 2.a) to indicate any changes to the source prominently. Also, the modified program itself may only be distributed under the GPL.
Thus, the GPL itself makes it hard to cheat by block-copy without being noticed.
The real problem is that of someone consulting the "answer" while they write their own independent implementation, and there is little you can do to stop it.
Here's how I'd word the paragraph, personally.
To view code I have written, you may
visit my off-campus website.
Some of these programs were written as solutions for problems presented in class at the University of Delaware. It is therefore worth noting to current University of Delaware students that
using or consulting these programs in order to complete your assignments may be in violation of the
University of Deleware's Academic dishonesy policy.
I think the issue is the time-thresholds involved. Older monitors could burn in relatively quickly. Modern displays take a bit more "persistence" to burn in a specific pattern. Old projection TVs were really bad about burnin -- playing video games with "unnaturally white" objects onscreen could scar the CRTs. (That probably had something to do with the light-levels involved in order to be able to project the image.)
The problem hasn't gone away, but I'd say it's lower magnitude than it used to be. For instance, you don't see too many modern arcade games with "INSERT COIN" burned into the CRT tube any more, but it was very common 20 years ago. How many Win9x and later machines have you seen with the Start bar burned in?
It sounds like the problem list operators face is that Person A may forge a subscription request from Person B (say, as a prank). The confirmation letter that Person B receives but did not request is considered actionable spam. This places the list operator in a pickle.
The irony is that the confirmation letter is the primary mechanism to prevent pranksters from signing up thier targets en masse to a series of mailing lists.
I think the intention is to punish those who send confirmation messages for which there is no outside subscription request as spam. That is, remove prankster Person A from the above sequence, and insert the list operator in their place. Such a scenario is similar to the "you've been preapproved to receive this blah blah; call 900-xxx-xxxx with confirmation number" type of snail-mail spam we receive.
I would think the list operator could probably shed liability in the prankster case by claiming that the subscribe request was made under false pretenses, and so the list operator himself was defrauded. Thus, the spam is the liability of the person sending the forged subscription request and not the list operator or the recipient.
While it's true you shouldn't plan on bonuses, more and more high-tech companies are giving bonuses in lieu of raises. It seems really nice, when times are good, because you get these huge bonuses--you think you're something really special. And after a few years, it takes quite a bit of fiscal discipline for those bonuses to not make their way into your usual budget. Then, the market goes in the tank, and you're suddenly making half of what you used to since the company can't afford the bonuses due to market conditions.
And yes, I'm taking home half of what I did last year, even though my base salary didn't change. I'm not hurting, though -- most of my expenses are recreational and non-recurring. Both my cars are paid for too, which helps a lot.
On the plus side, bonuses provide compensation flexibility for the company such that it's easier for the company to adjust its personnel expenses without actually letting as many people go. I've been through a few Reductions In Force, and it seems that the primary targets are the bottom 10% performance-wise, and top 10% salary-wise (at least among the non-upper-management staff). Thus, if your base wage stays lower, it's easier for the company to not lay you off -- just pay you less. Moral of the story: Sock those bonuses away for a rainy day.
You have the wrong wife.:-) Mine got excited about finding a Commodore 64 at a Hamfest, and has a dual boot PC (Linux/Win98), dual boot Mac (MacOS 9/MacOS X) and dual boot iBook (MacOS 9/Linux)...
Oh, and being a geologist, she hates diamonds.:-)
Now if I could only pull her away from EverQuest.:-/
Thanks for the clarifications! I guess from your description, the Feynman's FSM Firing Squad can be considered a 1-D array of cellular automata, correct?
BTW, by "specialized", I meant that the notion of FSM embodied in a cellular automata is specialized over the more general notion of a state machine -- that is, cellular automata are a specialized form of FSM.
I didn't mean to imply by the word "specialized" that different cells have different programs--although, you could provide for that in a single program by providing disconnected subsets of the states in your single "program" (which also solves the "MIMD initialization problem" you mention offhand).
Here's an example of what I mean by disconnected subsets. Suppose each cell has 6 states labelled A through F. Suppose A, B and C all have rules for transitioning among A, B and C. Similarly, suppose D, E, F have rules for transitioning among D, E and F. You can't reach A, B, or C if you start in D, E, or F, and vice-versa. If all your cells have programs of this form, then which subprogram a given cell 'runs' depends only on which subset of the allowed statespace you allocate that cell's starting state from.
As I understand it (someone correct me if I'm wrong), cellular automata have an FSM at their heart. Their inputs are determined spatially -- eg. their neighbors outputs serve as inputs. Also, I believe the output of cellular automata's FSM is generally of the form "alive" or "dead". Think, for instance, of Conway's "Life", or the various "Rule XX" that Wolfram examined that lead to him writing ANKOS.
So, in other words, cellular automata are a bunch of specialized FSMs.
More arbitrary FSMs can take more arbitrary inputs and produce more arbitrary outputs (or sets of outputs).
Slashdot Mirror
Restrictive, insofar as it's places more restrictions on you than a work that is not protected by copyright. GPL relaxes the restrictions that copyright places on you. It does not eliminate them or add new restrictions. It does trade absolute restrictions (Thou shalt not duplicate) for conditional restrictions (Thou shalt not duplicate, unless you also provide source under GPL).
So, to recap, the GPL is a restrictive license to the extent that relaxing a restriction without eliminating completely is itself a restriction.
--JoeMicrosoft licenses also limit the things you can do around the code, such as (in various EULAs) report benchmark numbers, write reviews, etc.
--JoeRelying on slashdot to get washing machine mod ideas? Ack no! All you'll get are neon light inserts, acrylic panels, and maybe a Matrix Orbital display w/ backlit keypad to replace the dials on the top.
How would theming your washing machine help brew beer?
--JoeIf you read the article, it says he ran the machine 20 times or so w/ just water to remove the soap residue.
As long as you can download the source code or get it on a floppy/CD/whatever, it's all good.
GPL says that any redistribution of a GPL'd work or derivative must include an offer of source. It doesn't say you need to make binaries available or installable ISOs available.
--JoeIt's probably more like somebody happened to be looking for the ROM for some other purpose, noticed a bunch of symbols that looked familiar, and started poking around the Linux kernel to compare. (They may have even seen "Linux" or similar strings.) In this case, the affected code is PCI setup code, so maybe the person who noticed was debugging some problem on their particular system that uses this product, or were developing a PCI driver.
--JoeUnder UNIX, use "file foo " to try to determine the type of contents in foo. It uses pattern information in the file /etc/magic to make its guess.
--JoeI'm obviously on crack. Columbia flew 28 missions.
Actually, the odds are much better than that. Columbia flew well over 100 missions before we lost it, and we've had only two shuttle disasters. I seem to recall hearing that the odds are estimated at about 1 in 250 that a given shuttle launch will result in loss of equipment or personnel. Still not great odds, but nearly an order of magnitude better than 1 in 50.
What gets me is that even if NASA knew reentry was going to destroy the shuttle after reviewing the launch videos, there wasn't much (anything?) they could really do to save the astronauts. Rather, they get to float around for 2 weeks waiting to die. That's really neat. :-P
--Joe --JoeI think the parent was thinking of Q, not Bond.
Threaded trees are kinda cool, but they do have some additional storage associated with them. Specifically, you have to have some way of distinguishing between a pointer to a child and a pointer back up into the tree. Also, how do you manage to get a post-order traversal out of one?
Back to the topic: Once you make the translation from code-recursion to data-recursion, you also open yourself up to other neat transformations. For instance, the difference between a depth-first search and a breadth-first search is simply the difference between a stack and a queue. If you write your code appropriately, you can switch between one or the other painlessly. And if you use a priority queue, you can have a loop which morphs between the two or gives you something in-between, just by changing the priority function. (I find this last structure useful for maze generation.)
--JoeSo you trade code recursive for data recursive. It's still recursive. Often, data recursive code is more complex to get right, too.
As the parent post noted, some algorithms are inherently recursive. Unfortunately, many teaching examples given to beginners don't need recursion.
Iterative tasks can be made recursive, but there's little point. For example, consider the classic "factorial()" and "fibonacci()" examples, which are really just iterative examples in which previous iterations feed future iterations. Although these tasks are "recursive" in the mathematical sense -- later elements are defined in terms of previous elements -- the computation itself is iterative.
Tasks which "fork", such as depth-first traversals of complicated structures, aren't simple iterative processes, since there is an implicit "go back" step that rewinds to a previous context where the fork occurs. You can translate these into a loop that lacks recursive function calls, but only if you sprout a data stack to keep track of these forking points. You've really just traded one stack for another and have not transformed the problem.
--JoeAnd how does that help me log in with SSH, SCP, telnet or FTP? The webpage would need to give me some sort of textual authentication token as part of mastering a series of clicks. I would then give telnet/ftp/ssh/scp the textual authentication token in order to log in.
And where would I provide that token to these programs? At the password prompt.
I guess one thing you could do with this is make the passwords single-use only.
--JoeSo have I. Mozilla 1.2 actually does this.
--JoeI suppose technically so. A slight change in wording is all that's necessary. The GPL requires you (see paragraph 2.a) to indicate any changes to the source prominently. Also, the modified program itself may only be distributed under the GPL.
Thus, the GPL itself makes it hard to cheat by block-copy without being noticed. The real problem is that of someone consulting the "answer" while they write their own independent implementation, and there is little you can do to stop it.
Here's how I'd word the paragraph, personally.
--JoeNeither does CSS. Yet, look at all the stink surrounding DeCSS.
The DMCA criminalizes the circumvention of access controls, not copy controls. The encryption in brunson's sig servse as an access control.
--JoeI think the issue is the time-thresholds involved. Older monitors could burn in relatively quickly. Modern displays take a bit more "persistence" to burn in a specific pattern. Old projection TVs were really bad about burnin -- playing video games with "unnaturally white" objects onscreen could scar the CRTs. (That probably had something to do with the light-levels involved in order to be able to project the image.)
The problem hasn't gone away, but I'd say it's lower magnitude than it used to be. For instance, you don't see too many modern arcade games with "INSERT COIN" burned into the CRT tube any more, but it was very common 20 years ago. How many Win9x and later machines have you seen with the Start bar burned in?
--JoeI thought that was called 'astroturfing'. The viral part I guess is in the "pass it on!" aspect of it?
--JoeIt sounds like the problem list operators face is that Person A may forge a subscription request from Person B (say, as a prank). The confirmation letter that Person B receives but did not request is considered actionable spam. This places the list operator in a pickle.
The irony is that the confirmation letter is the primary mechanism to prevent pranksters from signing up thier targets en masse to a series of mailing lists.
I think the intention is to punish those who send confirmation messages for which there is no outside subscription request as spam. That is, remove prankster Person A from the above sequence, and insert the list operator in their place. Such a scenario is similar to the "you've been preapproved to receive this blah blah; call 900-xxx-xxxx with confirmation number" type of snail-mail spam we receive.
I would think the list operator could probably shed liability in the prankster case by claiming that the subscribe request was made under false pretenses, and so the list operator himself was defrauded. Thus, the spam is the liability of the person sending the forged subscription request and not the list operator or the recipient.
--JoeWhile it's true you shouldn't plan on bonuses, more and more high-tech companies are giving bonuses in lieu of raises. It seems really nice, when times are good, because you get these huge bonuses--you think you're something really special. And after a few years, it takes quite a bit of fiscal discipline for those bonuses to not make their way into your usual budget. Then, the market goes in the tank, and you're suddenly making half of what you used to since the company can't afford the bonuses due to market conditions.
And yes, I'm taking home half of what I did last year, even though my base salary didn't change. I'm not hurting, though -- most of my expenses are recreational and non-recurring. Both my cars are paid for too, which helps a lot.
On the plus side, bonuses provide compensation flexibility for the company such that it's easier for the company to adjust its personnel expenses without actually letting as many people go. I've been through a few Reductions In Force, and it seems that the primary targets are the bottom 10% performance-wise, and top 10% salary-wise (at least among the non-upper-management staff). Thus, if your base wage stays lower, it's easier for the company to not lay you off -- just pay you less. Moral of the story: Sock those bonuses away for a rainy day.
--JoeYou have the wrong wife. :-) Mine got excited about finding a Commodore 64 at a Hamfest, and has a dual boot PC (Linux/Win98), dual boot Mac (MacOS 9/MacOS X) and dual boot iBook (MacOS 9/Linux)...
Oh, and being a geologist, she hates diamonds. :-)
Now if I could only pull her away from EverQuest. :-/
--JoeThanks for the clarifications! I guess from your description, the Feynman's FSM Firing Squad can be considered a 1-D array of cellular automata, correct?
BTW, by "specialized", I meant that the notion of FSM embodied in a cellular automata is specialized over the more general notion of a state machine -- that is, cellular automata are a specialized form of FSM. I didn't mean to imply by the word "specialized" that different cells have different programs--although, you could provide for that in a single program by providing disconnected subsets of the states in your single "program" (which also solves the "MIMD initialization problem" you mention offhand).
Here's an example of what I mean by disconnected subsets. Suppose each cell has 6 states labelled A through F. Suppose A, B and C all have rules for transitioning among A, B and C. Similarly, suppose D, E, F have rules for transitioning among D, E and F. You can't reach A, B, or C if you start in D, E, or F, and vice-versa. If all your cells have programs of this form, then which subprogram a given cell 'runs' depends only on which subset of the allowed statespace you allocate that cell's starting state from.
--JoeAs I understand it (someone correct me if I'm wrong), cellular automata have an FSM at their heart. Their inputs are determined spatially -- eg. their neighbors outputs serve as inputs. Also, I believe the output of cellular automata's FSM is generally of the form "alive" or "dead". Think, for instance, of Conway's "Life", or the various "Rule XX" that Wolfram examined that lead to him writing ANKOS.
So, in other words, cellular automata are a bunch of specialized FSMs.
More arbitrary FSMs can take more arbitrary inputs and produce more arbitrary outputs (or sets of outputs).
--JoeVAXaholic -- I recognize that nick from about 10 years ago. It's a unique enough nick that I don't feel silly asking: What school did you go to?
--Joe